Bunch-Shape Measurements at PSI s High Power Cyclotrons and Proton Beam Lines

Transcription

1 Bunch-Shape Measurements at PSI s High Power Cyclotrons and Proton Beam Lines Rudolf Dölling, Paul Scherrer Institut, CH-5232 Villigen-PSI technique - measurement locations, measurement principle - setup of detectors and timing&other electronics - measurement and evaluation procedure, corrections, software results - on beam parameters - on the methods performance/problems - on wire probe performance eventual next steps relation to beam dynamics simulations and machine development

2 measurement locations technique

3 measurement principle technique

4 measurement principle technique

5 detector setup at beam lines with several wire orientations several 2D projections of 3D density distribution carbon wires Ø33 um with current read out (schematic, seen in beam direction, the broader printed wire ends are closer to the beholder) scintillator PMT Mumetal shield technique

6 detector setup at Injector 2 scintillator wire replacement technique

7 detector setup at Ring cyclotron shielding against extraction elements EEC, FM scintillator (moves with wire) technique

8 electronics - timing modules (NIM) - relays - high voltage - wire current readout (logarithmic ampl.) - motor drivers technique

9 setting the PMT beam energy (at beam lines ) discrimination in order to only accept a single species: - instead of adjusting the discriminator level, the PMT voltage gain pulse height is varied - precision is needed (1V steps or better) - stability is needed (to prevent walk) - in some locations to be checked weakly (degradation of scintillator due to radiation) (wire at fixed position in beam, not same integration time) 72 MeV (small aperture) "clean" separation all elastically scattered usable p elastic inelastic 72 MeV (large aperture) only 40% of all elastically scattered usable p,n,d,? He? 590 MeV (not stopped) no discrete energies only a few slower particles usable long measurement duration technique

10 setting the PMT beam energy (Injector 2) Bunch center energy changes from turn to turn PMT voltage to be varied with probe position. Some error introduced by assumptions on how local beam energy increases with radius: - increase per turn, linear with bunch center radius (betatron oscillations introduce error) - same energy all over a bunch (not linear with actual radius, effect of space charge induced vortex motion?) beam dynamic simulations needed for information (Pulse-height resolution not good enough to measure energy differences in bunch.) voltage adapted for (still too high) contour levels every 10% and at 1% and 0.1% (10%-level at border between cyan and light blue) halo over-emphasized MeV (new detector) technique

11 corrections at beam energy (Injector 2) effects consequences evtl. corrections - distance wire detector changes shifts TOF of elastically scattered proton geometric correction shifts solid angle to detector aperture geometric correction - systematic variation of beam energy shifts TOF of elastically scattered proton geometric correction* * with radius (in bunch and shifts PMT pulse height (walk) PMT voltage adapted at meas. from turn to turn) shifts scattering cross section empiric correction* * - time resolution of measurement elongates crude correction * with assumptions on energy variation * can be accounted for by including scattering and transport to detector in beam dynamics simulation (predict histogram) More issues, all elongating, hardly to correct for: - beam energy spread at each radius spreads TOF * - detector aperture allows range of scattering angles spreads energy and cross section * - quantum efficiency/gain changes over PMT surface affects PMT pulse height (walk) - light collection efficiency dependent on impact position affects PMT pulse height (walk) affects TOF of light & PMT transfer time And - PMT base line distortion (by EMV or background radiation) systematically/statistically affects discrimination significantly more complicated than e.g. wire monitor evaluation technique

12 measurement software measurement modes (can be chosen for every wire) - 2D projection of bunch shape (standard): slice time-structure measured at a serious of wire positions ~6 minutes/full projection ~30 min in Ring cyclotron (smaller aperture, not stopped) - check of PMT voltage: slice time-structure measured for several PMT voltages (at fixed wire position) - check of time resolution: as above but coincidence signal instead of reference signal functionality - sets relays - proposes useful voltage and position ranges for all locations & measurement modes - steers drives, starts/stops/reads MCA (waits if beam is missing) - monitores PMT base current for over-current condition - logs ~500 machine parameters (settings, losses) plus wire current, plus PMT voltage & base current at each wire position (min/max/av) ( test case for simulations ) - (some machine interlock levels has still to be increased by hand to allow for increased losses from wire) - gives progress information - still not a standard application technique

13 evaluation software measurement settings - starts with useful time and position ranges for all locations & measurement modes - performs corrections (configurable) - shows 1 logged machine parameter (out of ~500) - writes data to files evaluation & display settings results input for simulation technique

14 technique - measurement locations, measurement principle - setup of detectors and timing&other electronics - measurement and evaluation procedure, corrections, software results - on beam parameters - on the methods performance/problems - on wire probe performance eventual next steps relation to beam dynamics simulations and machine development

15 beam at Injector 2 last turns MeV production beam 2200 ua - Three separate scans with correspondingly adapted PMT voltage ramps. - A relative phase slip of ~9 builds up over the last 11 turns. results

16 beam at Injector 2 last turns condensed to bunch parameters of many turns at 2200 ua extracted from the three scans (plus two repetitive scans, one with increased PMT voltage) individually derived from a combination of two scans (lines only to guide the eyes) results

17 beam after at Injector 2 production beam 2200 ua results

18 performance: artefacts? possible sources of artificial counts: - background radiation (especially difficult when correlated i.e. created by loss generated by the wire) - coupling of stray RF fields to measurement cable (correlated) - reflections in timing circuit even in a "quiet environment" it is difficult to judge what is an artefact - transversal tails are presumably real - longitudinal tails may eventually be artefacts production beam 2200 ua behind Injector 2 "quiet environment" results

19 performance: dynamic range 2 na 47 pa production beam 2200 ua last turns in Injector 2 "quiet environment" results

20 performance: dynamic range an example of a low dynamic range: raw data filtered Dölling, HB2010 production beam 2200 ua in Ring cyclotron at high beam loss not a "quiet environment (and a degraded scintillator) results

21 "old" setup performance: time resolution "new" setup results

22 performance: time resolution correction of resolution (assumed to be 13.5 ps) by quadratic subtraction: this comes to its limits! This is at 72 MeV! At 590 MeV it is worse but still good enough. Dölling, HB2012 results beam 50 ua after Injector 2

23 reference for wire probes/monitors with current reading wire current is measured production beam 2200 ua last turns in Injector 2 carbon wire Ø33 um "quiet environment" Thermionic emission dominates current signal of slowly moved wire probe in narrow beam (@72 MeV). Can be suppressed by positive wire bias. (Simulations: will be <1% of regular signal if wire speed >1 m/s) Stray particles limit dynamic range of wire probe (depends on environment). When thermionic emission is not developed, 0 V bias gives the best result (in this environment). Bunch shape measurement is clearly superior (but slow). results

24 technique - measurement locations, measurement principle - setup of detectors and timing&other electronics - measurement and evaluation procedure, corrections, software results - on beam parameters - on the methods performance/problems - on wire probe performance eventual next steps relation to beam dynamics simulations and machine development

25 improvements at 590 MeV longer drift needed for full separation? E discriminator acceptance window better rate (LED acceptance level only a few slower particles usable) no discrete energies coincidence measurement second detector after ~0.5 m further path has also to be passed for acceptance better immunity against background radiation better dynamic range eventually an additional probe at the last turns of the Ring cyclotron eventual next steps

26 beam energy measurement at 72 MeV The energy of the beam after the Injector cyclotron can be determined to ~1e-3 by making the distance detector-wire variable (by setting the detector on a motorized feedthrough) and taking the time spectra at two different distances. Eventually it is possible to unfold some details of the beam energy distribution. eventual next steps

27 technique - measurement locations, measurement principle - setup of detectors and timing&other electronics - measurement and evaluation procedure, corrections, software results - on beam parameters - on the methods performance/problems - on wire probe performance eventual next steps relation to beam dynamics simulations and machine development

28 my personal view on the future development of our machine Yang, Bi, Wei, Zhang, Adelmann,... relation to beam dynamics simulations and machine development

29 Bunch-Shape Measurements at PSI s High Power Cyclotrons and Proton Beam Lines Rudolf Dölling, Paul Scherrer Institut, CH-5232 Villigen-PSI technique - measurement locations, measurement principle - setup of detectors and timing&other electronics - measurement and evaluation procedure, corrections, software results - on beam parameters - on the methods performance/problems - on wire probe performance eventual next steps relation to beam dynamics simulations and machine development Thanks for listening!

30 back-up slide: beam at Injector 2 last turns condensed to bunch parameters of last turn at varied current extracted from the three scans (plus two repetitive scans, one with increased PMT voltage) individually (parameters logged during measurement) derived from a combination of two scans back-up slides

31 back-up slide: "super-buncher" idea: restoring short beam at entrance of Ring cyclotron roll-up there (?) (layout based on 1D bunching simulation) preliminaty tests: full voltage / envisaged operation not possible yet due to increased losses probable explanation: difficult to match beam and halo M. Humbel et al., this conference Schmelzbach et al., EPAC06 Schmelzbach et al., EPAC06 J. Yang et al., HB2008 back-up slides

32 possible strategy: understanding beam losses in detail back-up slide: simulations where (at low energies) to cut and how to match the halo ( collimation system ) controlled beam tails, matching of beam & halo lower losses this needs detailed simulations detailed input from diagnostics bunch-shape measurement, halo measurement what precision of measurement & simulation is needed? (maybe less than anticipated at first glance) encouraging steps done (OPAL code includes space charge, fields, scattering, optimisation, not neutralisation) still much to do (put many details to input file: collimators, trim coils, measured profiles,... space-charge neutralisation at 0.87 MeV?) and still far from full description or prediction will it work? it is essential for further development of the machine back-up slides